Reliable motion prediction of surrounding vehicles is the key to safe and efficient driving of autonomous vehicles, especially at urban intersections with complex traffic environments. This study models driving intentions and future driving speeds at urban intersections and improves model prediction performance by considering traffic environment characteristics. Key feature parameters including environmental characteristics are first extracted through driving behavior analysis and existing research experience. Then models with different input combinations are constructed to explore the effectiveness of different factors in predicting driving intention and future speed. In particular, in vehicle speed modeling, a target detection algorithm is used to identify traffic participants. Based on the identified traffic participant and vehicle position information, a new method for speed prediction that can reflect the dynamic interaction characteristics between the driver and the traffic environment is proposed. Models are trained and tested using natural driving data from China. Finally, the models with the simplest input and the best effect are determined. The driving intention recognition model can accurately predict the driving maneuvers of straight-ahead, stopping, turning left and right 4 seconds before reaching the intersection. The speed prediction model can significantly improve the speed prediction accuracy, and shows stronger robustness and adaptability than existing models. This research provides important technical support for developing intelligent driving systems suitable for complex urban traffic environments.

BACKGROUND: Histological feature representation is advantageous for computer aided diagnosis (CAD) and disease classification when using predictive techniques based on machine learning. Explicit feature representations in computer tissue models can assist explainability of machine learning predictions. Different approaches to feature representation within digital tissue images have been proposed. Cell-graphs have been demonstrated to provide precise and general constructs that can model both low- and high-level features. The basement membrane is high-level tissue architecture, and interactions across the basement membrane are involved in multiple disease processes. Thus, the basement membrane is an important histological feature to study from a cell-graph and machine learning perspective. RESULTS: We present a two stage machine learning pipeline for generating a cell-graph from a digital H &E stained tissue image. Using a combination of convolutional neural networks for visual analysis and graph neural networks exploiting node and edge labels for topological analysis, the pipeline is shown to predict both low- and high-level histological features in oral mucosal tissue with good accuracy. CONCLUSIONS: Convolutional and graph neural networks are complementary technologies for learning, representing and predicting local and global histological features employing node and edge labels. Their combination is potentially widely applicable in histopathology image analysis and can enhance explainability in CAD tools for disease prediction.

Background: Program similarity is a fundamental concept, central to the solution of software engineering tasks such as software plagiarism, clone identification, code refactoring and code search. Accurate similarity estimation between programs requires an in-depth understanding of their structure, semantics and flow. A control flow graph (CFG), is a graphical representation of a program which captures its logical control flow and hence its semantics. A common approach is to estimate program similarity by analysing CFGs using graph similarity measures, e.g. graph edit distance (GED). However, graph edit distance is an NP-hard problem and computationally expensive, making the application of graph similarity techniques to complex software programs impractical. Aim: This study intends to examine the effectiveness of graph neural networks to estimate program similarity, by analysing the associated control flow graphs. Method: We introduce funcGNN1, which is a graph neural network trained on labeled CFG pairs to predict the GED between unseen program pairs by utilizing an effective embedding vector. To our knowledge, this is the first time graph neural networks have been applied on labeled CFGs for estimating the similarity between highlevel language programs. Results: We demonstrate the effectiveness of funcGNN to estimate the GED between programs and our experimental analysis demonstrates how it achieves a lower error rate (1.94 x10-3), with faster (23 times faster than the quickest traditional GED approximation method) and better scalability compared with state of the art methods. Conclusion: funcGNN posses the inductive learning ability to infer program structure and generalise to unseen programs. The graph embedding of a program proposed by our methodology could be applied to several related software engineering problems (such as code plagiarism and clone identification) thus opening multiple research directions.

We give an introduction to the LBTest tool which implements learning-based testing for reactive systems. It makes use of incremental learning and model checking algorithms to automate: i) test case generation, ii) test execution and iii) test verdict construction. The paper illustrates the tool by means of a pedagogical case study, to enable the user to setup and learn the tool quickly. We provide a usability exercise to support tool evaluation.

We consider black-box testing of functional correctness as a special case of a satisfiability or constraint solving problem. We introduce a general method for solving this problem based on function approximation. We then describe some practical results obtained for an automated testing algorithm using approximation by piecewise polynomial functions.